Duckitt and Foster (1991) note that in the period of the late 1970s to the early 1980s, South African research on racial attitudes was plagued with three core damaging features. Firstly, few studies were conducted at that time that examined the symbolic racism argument in the country. Secondly, research that had been conducted had often utilised inappropriate measures that reflected American indices of racism. Lastly, the authors note the consistent
sample bias of these studies. A number of years later, it appears that two of these concerns
•
still plague South African racial attitude research. Although more research on the relevance of symbolic racism in the country has been conducted since that time, on the whole present studies still reflect the use of inadequate measures and a sample bias; these two points are considered next.
3.3.1 The use of inadequate measures
Since few measures have been developed for the South African context, research on racism reflects a distinct trend in using modified international measures of prejudice and racial attitudes. For example, in a past study, Heaven and Moerdyk (1977, in Foster, 1991) used an adapted anti-black scale to assess South African attitudes; however, this scale was originally developed by Ray (1976) to measure attitudes of aborigines in Australia (Foster, 1991).
Since then the scale has been used on a number of occasions in subsequent studies examining a number of different racial issues in South Africa (for example, Heaven, 1978; Heaven &
Stones, 1979, in Foster, 1991; Duckitt, 1990; 1991).
A distinct problem with the use of such scales in the South African context is that they have questionable content and face validity since it is unclear the extent to which the content constituting the scales relates, and appears to relate, to local people. These scales have been developed on the basis of international conceptualisations and theories of racism; their use in South African studies, therefore, may compromise the internal validity of local studies to some degree. Although past studies have modified items constituting the scales; for example, Ray's anti-black scale was found reliable for South African people by modifying items (Heaven & Stones, 1979, in Foster, 1991), the items themselves still reflect the international domain of racial content. As a result, South African samples from previous studies have been presented with racial items that reflect, both in content and on face value, international racial foci. This points to another distinct problem, which is that there has been little attempt to formally establish the domain of content constituting South African racism. As a
consequence of this, there has been no real basis to compare and cross validate the content constituting local racial opinions with those of international ones.
Although Duckitt's subtle racism scale marks one of the few exceptions which represent a locally developed measure, in light of very recent racial attitude item development (for example, Sears & Henry, 2003), this scale appears outdated, with many of these items reflecting early symbolic content of the 1980's. For example, one item reads, "Given the same education and opportunities, blacks should be able to perform as well as whites in any field", whilst another reads, "Itis almost certainly best for all concerned that interracial marriages not be allowed". According to Sears and Henry (2003), by recent standards, such items may not be sufficiently sophisticated to divulge the difficult style of racism
characterising contemporary society. Sophisticated racists today would probably not deny that blacks could perform as equally as well as whites given the same opportunities.
Similarly, it would be far too crude for symbolic racists to agree that inter-racial marriages should not be permitted. These items, therefore, may be inadequate in light ofthe recent conceptualisations of racism. Although Duckitt and Mphuthing's (1998) revised version of the scale reflects items incorporating more recent South African content, a number of significant events have occurred since 1998 influencing race relations and attitudes. In this respect, the context necessitates an updated measure that reflects these issues.
In sum, South African research on racial attitudes could benefit from establishing some of the topics characterising South African racism and then using these to develop a revised, more relevant measure for its people. This will help ensure both the content and face validity of future measures and hence, enhance the overall internal and external validity of future studies.
3.3.2 Sample bias
Another damaging feature of racial prejudice studies in general, is that they have, and continue, to predominantly focus on white prejudice. Shelton (2000) outlines a number of reasons for this bias in the United States, but suggests one of the prime reasons may be on account of the fact that, historically, the United States has focussed on racism as a white problem and hence, has focus sed on the issues of prejudice from a white perspective. Shelton outlines two serious implications of this that; first, whites are portrayed as "active perceivers"
(p. 374) and blacks as "passive targets" (p. 374) of racial prejudice studies with no real attention given to understanding how blacks themselves influence inter-group racial
dynamics. Second, and as a consequence of this first point, research has failed to examine racial prejudice from what Shelton terms, "a true interpersonal perspective in terms of the dynamics of cross-racial social interactions (p. 375). She suggests instead, that studies on racial prejudice need to "examine not only how Blacks respond to whites' behaviours, but also [emphasise] Blacks' behaviours toward Whites as an equally important issue" (p. 388).
In doing this, studies can then begin to understand prejudice from the perspective of blacks and not only whites.
In a similar vein to the United States then, South Africa too has demonstrated favour in studying the views of whites. In light of South Africa's past, however, this bias may be somewhat justified for different reasons. There are, however, grounds, particularly in
contemporary South Africa, to examine the views of other race groups. South Africa has also undergone enormous change and the country reflects significantly greater racial integration than former years. In accordance with Shelton's (2000) argument, it seems imperative that South African studies begin emphasising the role of blacks in racial prejudice studies as equally important; this will not only emphasise blacks as active beings in the research process, but will also provide the perspective of blacks needed in understanding the interpersonal aspect of racial prejudice.
In terms of sample bias, South African studies are also subject to criticism on the basis of drawing conclusions from student samples. The use of students as research participants may be attributed to their accessibility and convenience as a source of gathering information.
However, students often reflect liberal views, which may be unrepresentative of the general population of South Africa.
SECTION B: RESEARCH, FINDINGS, AND CONCLUSIONS
CHAPTER 4
AIMS AND PROCEDURE OF RESEARCH